Scenario:
atomic_t start, end;
current is the only task in wq;
consumer does: wait_even_interruptible(wq, start != end)
producer does: end++; wakeup_interruptible(wq)
Is it possible that the task in wq will be scheduled twice,
once since end >= start (mod something) and again since,
wakeup_interruptible was called?
--
Luben